1G4 
PROFESSOR W. N. HARTLEY ON 
lilies due to the flame spectra are marked out bj the appearance of minute dots 
Where the insensitive portion of the film occurs, strong lines are easily seen on the 
continuous linear spectrum in consequence of the slit being slightly widened for a 
minute portion of its length, so that the effect caused by want of sensitiveness in the 
silver salts is diminished. 
It is a little difficult to read the measurements and describe the spectra at the 
same time, hence enlargements were made upon which the measurements were 
recorded as they were read off. Another convenient plan was to adjust the scale to 
the photograph and take an enlargement therefrom at once, so that prints from the 
same give approximately their own measurements. Only those measurements are 
exact which are exactly at the centre of the photograjDhic lens, even when the scale 
is precisely adjusted to the photograph, so, for instance, that the 20th division was 
exactly at the sodium line. In cases where the lines wei’e not newly discovered, and 
it was only necessary to identify them, nothing more was required. New lines and 
bands were measured by a micrometer screw with a pitch of 100 threads to the inch, 
and a wheel head divided into 100 parts. The screw carries a nut on which a micro¬ 
scope, magnifying 10 diameters, is fixed, by which arrangement it is easy to measure 
to r^rToo th of an inch, and, where desirable, to rooTbiTof^^- This instrument was 
made by Mr. A. Hilger, of London. Each measurement was recorded at the time 
by writing on an enlarged print of the same photograph. 
2'he Spectrum seen ivhen siqyports of Kyanite alone are heated in the 
Oxy-hydrogen Flame. 
Just as in the ordinary use of the spectroscope we are prepared to see the lines of 
sodium,and under certain circumstances the bands peculiar to carbon, so in these photo¬ 
graphs, the sodium lines and the strongest groups of lines belonging to the emission 
spectrum of water vapour, are also always jiresent. In addition to these, the kyanite 
yields the red line of lithium, wfliich is no inconvenience, but a positive advantage, as 
it serves to indicate where the spectrum commences, and from which point measure¬ 
ments may be made. 
The Extent and Character of the Spectra observed. 
Although the apparatus is capable of photographing on one plate rays lying 
betwmen wave-lengths 6708 of lithium in the red and 2194 in the ultra-violet, 
nevertheless the flame spectra of a large majority of the metals and their compounds 
terminate somewhere about the ultra-violet emission spectrum of water. The first, 
* For several of tlie enlarged negatives made exactly to the same scale I am indebted to the kindness 
of mj^ friend, Pi-ofes.sor Alec Fraser, ■who devoted much of his o'wn valuable time to making negatives 
with as perfect a definition as possible, the prints troin which have greatly facilitated my work. 
